⚡ Python/JS Apps

Python/JS apps are the full-control way to build on Keboola. Where Streamlit gives you a quick, ready-made Python environment, Python/JS apps let you bring your own code — any Python web framework (Flask, FastAPI, Dash), a JavaScript frontend (React, Vue, plain JS), or both together. You build the experience; Keboola handles the hosting, access control, and data connectivity.

Why choose Python/JS

• Any framework, any UI. Full control over look, layout, and interactivity — not limited to a widget set. Build a polished, customer-facing product, not just an internal tool.
• Full-stack. A JavaScript/React frontend with a server-side backend (for example, an Express or FastAPI API) in one app.
• Server-side data access. Your Storage token stays on the server; the browser never sees it. Query Keboola Storage (via the Query Service / Storage API) from your backend and shape exactly the data your UI needs.
• Expose an API — or an MCP server. Because there’s a real backend, the app can expose an API (or an MCP server) so another agent or service can call it, not just human visitors.
• Your Git workflow. Develop in your own editor and Git account and sync the app to your project — production-grade, versioned development.

Python/JS or Streamlit?

• Reach for Python/JS when you want a rich or customer-facing app, a custom frontend, a backend/API, or full control. See below to start building.
• Reach for Streamlit when you want the fastest path for a simple internal data tool in pure Python.

For the bigger picture of how apps fit into Keboola, see What are Keboola apps.

How a Python/JS app runs

Keboola clones your repository, installs dependencies, starts your app using the process configuration you provide, and serves it behind a secure URL through an internal web server. You don’t manage servers, ports, or Docker — only your code and a small configuration folder.

How development works

The day-to-day loop, whichever way you build:

1. Code lives in a Git repository — yours, or a private Keboola-managed repo that Kai creates for the app (git.<stack>.keboola.com/keboola/app-<id>.git). The scaffold is a React frontend (App.tsx), server-side routes (index.ts), and a data-access helper (kbcQuery.ts) — see App structure.
2. Data access happens server-side. Your backend queries Storage (Storage API or real-time SQL via the Query Service) using the auto-injected KBC_TOKEN — the browser never sees the token; the frontend only talks to your own API routes. Environment variables and code patterns are in Reference → Data access.
3. Ship a change: push to the connected branch and hit Redeploy (Keboola re-clones, reinstalls, restarts) — or, if Kai built the app, just tell Kai what to change and approve its edit.
4. Debug on the app detail: the app’s page has Overview / Advanced Settings / All Runs / Terminal Logs / Drafts / Versions tabs — deployment state, run history, and terminal output are all there — and the deploy wizard controls backend size and the inactivity timeout. The app sleeps when idle and wakes on the next visit. When Kai edits a draft, the preview hot-reloads.

Build faster with AI Kit

The AI Kit gives your coding assistant (Claude Code, Cursor, Copilot) Keboola-specific skills — including the Data App Developer plugin with a validate → build → verify workflow for apps. When you create a Python/JS app, the dialog also offers Download Skill — a ready-made skill file teaching your assistant the correct app structure, deployment config, and Keboola APIs. Grab it there, download it directly, or install the full kit:

/plugin marketplace add keboola/ai-kit
/plugin install dataapp-developer

Build a Python/JS app

You can build one three ways:

• Build with Kai — describe it and let Kai generate the app. The fastest path. (Building Python/JS with Kai requires the Kai-in-E2B backend on your project; without it, Kai builds Streamlit. )
• Create the app manually — set the app up from the interface and point it at a Git repository that contains your code and configuration.
• Build locally — develop with your own tools and Git account, then sync to your project. The home of full-control development.

Building it yourself? When you create a Python/JS app, Keboola offers a downloadable AI skill (for Claude Code, Cursor, or Copilot) that teaches your assistant the correct app structure, deployment config, and Keboola APIs — see Develop with an AI coding tool.

Reference

Environment variables, the app scaffold, backend versions, and limits live in the Apps reference.